Deep learning based spectral extrapolation for dual‐source, dual‐energy x‐ray computed tomography

Abstract

Purpose: Data completion is commonly employed in dual-source, dual-energy computed tomography (CT) when physical or hardware constraints limit the field of view (FoV) covered by one of two imaging chains. Practically, dual-energy data completion is accomplished by estimating missing projection data based on the imaging chain with the full FoV and then by appropriately truncating the analytical reconstruction of the data with the smaller FoV. While this approach works well in many clinical applications, there are applications which would benefit from spectral contrast estimates over the larger FoV (spectral extrapolation)-e.g. model-based iterative reconstruction, contrast-enhanced abdominal imaging of large patients, interior tomography, and combined temporal and spectral imaging. Methods: To document the fidelity of spectral extrapolation and to prototype a deep learning algorithm to perform it, we assembled a data set of 50 dual-source, dual-energy abdominal x-ray CT scans (acquired at Duke University Medical Center with 5 Siemens Flash scanners; chain A: 50 cm FoV, 100 kV; chain B: 33 cm FoV, 140 kV + Sn; helical pitch: 0.8). Data sets were reconstructed using ReconCT (v14.1, Siemens Healthineers): 768 × 768 pixels per slice, 50 cm FoV, 0.75 mm slice thickness, "Dual-Energy-WFBP" reconstruction mode with dual-source data completion. A hybrid architecture consisting of a learned piecewise linear transfer function (PLTF) and a convolutional neural network (CNN) was trained using 40 scans (five scans reserved for validation, five for testing). The PLTF learned to map chain A spectral contrast to chain B spectral contrast voxel-wise, performing an image domain analog of dual-source data completion with approximate spectral reweighting. The CNN with its U-net structure then learned to improve the accuracy of chain B contrast estimates by copying chain A structural information, by encoding prior chain A, chain B contrast relationships, and by generalizing feature-contrast associations. Training was supervised, using data from within the 33-cm chain B FoV to optimize and assess network performance. Results: Extrapolation performance on the testing data confirmed our network's robustness and ability to generalize to unseen data from different patients, yielding maximum extrapolation errors of 26 HU following the PLTF and 7.5 HU following the CNN (averaged per target organ). Degradation of network performance when applied to a geometrically simple phantom confirmed our method's reliance on feature-contrast relationships in correctly inferring spectral contrast. Integrating our image domain spectral extrapolation network into a standard dual-source, dual-energy processing pipeline for Siemens Flash scanner data yielded spectral CT data with adequate fidelity for the generation of both 50 keV monochromatic images and material decomposition images over a 30-cm FoV for chain B when only 20 cm of chain B data were available for spectral extrapolation. Conclusions: Even with a moderate amount of training data, deep learning ...

Advances in micro-CT imaging of small animals

“…Finally, we note a trend in recent literature: estimating missing spectral information with DL priors. This has been demonstrated in the context of spectral extrapolation for dual-energy field-of-view extension [113], estimation of virtual monoenergetic images from single-energy data [114], and estimation of material maps form single-energy data [115]. Success in these applications speaks to the power of DL to model and enforce underlying relationships between image features and spectral contrast; however, significant work remains to understand the limitations and uncertainty inherent in these methods, with particular regard to unique or pathological data which may poorly represented.…”

Clinical Implementation of Dual-Energy CT for Gastrointestinal Imaging

“…Modern DECT scanners, however, use varying geometries, X-ray sources, detectors, and iterative or deep learning image reconstruction algorithms (Table 1) [7,8,10,12].…”

“…DECT radiation doses are almost comparable or even lower than a corresponding single-phase contrast-enhanced SECT acquisition, depending upon the scanner implementation, patient characteristics, and protocol [10,23,26]. Radiation dose optimization with DECT has been facilitated over the past few years by the advent of tin filtration, automatic dose modulation, and image denoising techniques such as iterative and, more recently, deep learning reconstruction algorithms [7,8,12,13]. The radiation dose optimization effort with DECT should always be approached in terms of whole examination radiation dose rather than from the perspective of a single acquisition only [10].…”

Untitled

“…Train and extrapolate models that are difficult to interpret at best by experts in the field. Neural networks can model complex nonlinear relationships, which has a benefit over humbler the methods used for the modelling, such as the Naïve Bayesian classifier or logistic regression [9,10]. The Supportive Vector Machines (SVM) is possibly one of the greatest influential algorithms used and applied for classification purposes today, where its analytical accuracy is of importance and consideration.…”